This application addresses the NIAAA Research Initiative to characterize the mechanisms of behavior change in drinking. Heavy alcohol use and alcohol use disorders (AUDs) peak between ages 18 and 25, representing a serious public health concern. It is estimated that roughly 50% of the risk for alcohol use disorders is genetic, and new technology has facilitated the identification of specific target genes that confer risk. In order to design effective early interventions, we must better understand the mechanisms through which target genes contribute to different patterns of drinking behavior. Involvement in multiple deviant behaviors (generalized deviance) and individual differences in alcohol response are established risk factors for later AUDs, and both are driven, at least in part, by genetic influences. The proposed study will address three major research aims: (1) Identify unique genetic associations with four different developmental trajectories of heavy drinking: Stable {High}, Late Increasing, Developmentally Limited, and Stable {Light}; (2) Identify genetic influences on individual differences in alcohol response in a laboratory-based, placebo-controlled alcohol challenge; and (3) Test alcohol response and generalized deviance as phenotypic mediators of the effects of five candidate genes on the four trajectories of heavy drinking. Our multi-method approach to these research aims is unique in four important respects. First, rather than rely on a cross-sectional sample, we will collect salivary DNA samples from an ethnically diverse prospective cohort (N = 1,060) who have previously reported on their alcohol use and other deviant behaviors (e.g., drug use, sexual risk-taking) across 10 waves of assessments from age 18 to 24. Second, we will capitalize on cutting-edge innovations in biotechnology that allow for cost- effective genotyping of candidate genes. We will focus on a panel of 63 polymorphisms in 5 candidate genes related to cholinergic (CHRM2), opioid (OPRM1), dopaminergic (DRD4), serotonergic (SLC64A), and GABA- ergic (GABRA2) neurotransmission, plus an array of Ancestry Informative Markers (AIMs) to identify ethnic ancestry and correct for population substructure. Third, 44% of the prospective sample will be tested in a placebo-controlled alcohol challenge to identify genetic influences on individual differences in response to the effects of alcohol. {Finally, we will assess mechanisms of genetic risk within a sophisticated structural equation modeling framework for longitudinal, multivariate data.} Thus, the proposed project will integrate 6 years of prospective behavioral data with new genotypying data and laboratory-assessed alcohol response endophenotypes. Results from the proposed project will facilitate early identification of individuals at high risk for developing an AUD and will contribute to the ultimate goal of developing more personalized targets of intervention.